Everyone has opinions about how tests should look. Naming conventions, structure patterns, which mocking library to use. The problem with AI coding assistants is that they have collected opinions for their learnings too, and they're usually not yours.
